EP2878678A1

EP2878678A1 - RNA-biomarkers for diagnosis of prostate cancer

Info

Publication number: EP2878678A1
Application number: EP13195377.0A
Authority: EP
Inventors: Friedemann Horn; Jörg Hackermüller; Sabina Christ; Kristin Reiche; Manfred Wirth; Michael Fröhner; Susanne FÜSSEL
Original assignee: Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Current assignee: Fraunhofer Gesellschaft zur Forderung der Angewandten Forschung eV
Priority date: 2013-12-02
Filing date: 2013-12-02
Publication date: 2015-06-03
Also published as: US20160304967A1; EP3077537B1; ES2784246T3; US10301683B2; WO2015082414A1; US20160298199A1; EP3077537A1; EP3077535B1; US20160304966A1; US10196696B2; US10196695B2; EP3077531A1; US20160304965A1; EP3077530A1; WO2015082416A1; EP3077531B1; US10287634B2; WO2015082418A1; WO2015082416A8; ES2732371T3

Abstract

The invention relates to the identification and selection of differentially expressed transcripts (biomarker) in tumour cells. Specific determination of the level of these biomarkers can be used for screening and diagnosis of prostate cancer. Clinical application of assays based on these biomarker help reduce the high number of false positives of current standard screening assays.

Description

Field of the invention

The present invention is in the field of biology and chemistry. In particular, the invention is in the field of molecular biology. More particular, the invention relates to the analysis of RNA transcripts. Most particularly, the invention is in the field of diagnosing prostate cancer.

Background

Prostate cancer is the most frequently diagnosed cancer in men. In 2012, the annual number of newly diagnosed prostate cancer cases was reported as approximately 240,000 cases in the United States and approximately 360,000 in the European Union, 68,000 of which in Germany. In the United States, lifetime risks for prostate cancer diagnosis and for dying of prostate cancer are currently estimated at 15.9% and 2.8%, respectively. Despite widespread screening for prostate cancer and major advances in the treatment of metastatic disease, prostate cancer remains the second most common cause of cancer death for men with over 250,000 deaths each year in the Western world.
Currently, testing of prostate-specific antigen (PSA) serum levels and the digital rectal examination represent the two major screening methods. Patients showing abnormal results usually are advised to have a prostate biopsy performed. This has however significant consequences. The lack of specificity of PSA screening which produces high numbers of false positives results in unnecessary prostate biopsies performed annually on millions of men worldwide (overdiagnosis). In addition, taking biopsies carries a substantial risk for infectious complications. Therefore, there is an urgent need for a more sensitive and specific diagnostic assay for early prostate cancer diagnosis to improve prostate cancer screening and to avoid the high numbers of unnecessarily taken prostate biopsies. The present invention addresses this problem by providing a set of biomarkers for the screening and diagnosis of prostate cancer.

Summary of the invention

By Next Generation Sequencing, transcripts were identified that are differentially expressed in 64 samples of prostate tumour and control tissues. The invention describes sets of these RNA biomarkers, which as of yet have not been described in the context of prostate cancer, for the use in diagnosis of prostate cancer.
The invention relates to a method for the diagnosis of prostate cancer, comprising the steps of analyzing a sample of a patient for the presence and/or level of nucleic acids selected from the group of SEQ ID NOs 1 to 42, wherein, if at least one of said nucleic acids is present and/or the level of at least one of said nucleic acid is above a threshold value, said sample is designated as prostate cancer positive. Of course also a fragment of said sequences may serve as biomarker. Such fragments may be as small as 10 nucleotides.
The invention relates to a probe or primer that hybridizes under stringent conditions to one of the said nucleic acids according to SEQ ID NO. 1 to 42.
The invention also relates to a probe or primer, wherein the probe or primer is specific for one of said sequences as selected from the group of SEQ ID NO. 1 to 42.
The invention relates to a nucleic acid from the group of SEQ ID NO. 1 to SEQ ID NO. 42 or the reverse complement thereof, or a nucleic acid that shares preferably at least 85%, 90%, 95% or 99% sequence identity with a nucleic acid according to any one of the nucleic acids according to SEQ ID NO. 1 to 42.
The invention relates to a nucleic acid selected from the group of SEQ ID NO. 1 to 42 for the diagnosis of prostate cancer.
The invention relates to a kit for the diagnosis of prostate cancer comprising at least one probe or primer as defined above.

Definitions

The following definitions are provided for specific terms, which are used in the application text.
The articles "a" and "an" are used herein to refer to one or to more than one (i.e. to at least one) of the grammatical object of the article. By way of example, "an element" means one element or more than one element. In contrast, "one" is used to refer to a single element.
The term, "analysing a sample for the presence and/or level of nucleic acids" or "specifically estimate levels of nucleic acids", as used herein, relates to the means and methods useful for assessing and quantifying the level of nucleic acids. One useful method is for instance quantitative reverse transcriptase PCR. Likewise, the level of RNA can also be analysed after amplification using spectrometric techniques at 260 and 280 nm respectively.
As used herein, the term "amplified", when applied to a nucleic acid sequence, refers to a process whereby one or more copies of a particular nucleic acid sequence is generated from a nucleic acid template sequence, preferably by the method of polymerase chain reaction. Other methods of amplification include, but are not limited to, ligase chain reaction (LCR), polynucleotide-specific based amplification (NSBA), or any other method known in the art.
The term "correlating", as used herein in reference to the use of diagnostic and prognostic marker(s), refers to comparing the presence or amount of the marker(s) in a patient to its presence or amount in persons known to suffer from, or known to be at risk of, a given condition. A marker level in a patient sample can be compared to a level known to be associated with a specific diagnosis.
As used herein, the term "diagnosis" refers to the identification of the disease (prostate cancer) at any stage of its development, and also includes the determination of predisposition of a subject to develop the disease.
As used herein, the term "fluorescent dye" refers to for example FAM (5-or 6-carboxyfluorescein), VIC, NED, Fluorescein, FITC, IRD-700/800, CY3, CY5, CY3.5, CY5.5, HEX, TET, TAMRA, JOE, ROX, BODIPY TMR, Oregon Green, Rhodamine Green, Rhodamine Red, Texas Red, Yakima Yellow, Alexa Fluor, PET and the like.
As used herein, "isolated" when used in reference to a nucleic acid means that a naturally occurring sequence has been removed from its normal cellular (e.g. chromosomal) environment or is synthesised in a non-natural environment (e.g. artificially synthesised). Thus, an "isolated" sequence may be in a cell-free solution or placed in a different cellular environment.
As used herein, a "kit" is a packaged combination optionally including instructions for use of the combination and/or other reactions and components for such use.
The term "level" in the context of the present invention relates to the concentration of a biomarker in a sample of a patient. Said sample from said patient is designated as prostate cancer positive if the level of a biomarker exceeds a threshold value.
As used herein, "nucleic acid(s)" or "nucleic acid molecule" generally refers to any ribonucleic acid or deoxyribonucleic acid, which may be unmodified or modified DNA. "Nucleic acids" include, without limitation, single- and double-stranded nucleic acids. As used herein, the term "nucleic acid(s)" also includes DNA as described above that contain one or more modified bases. Thus, DNA with backbones modified for stability or for other reasons are "nucleic acids". The term "nucleic acids" as it is used herein embraces such chemically, enzymatically or metabolically modified forms of nucleic acids, as well as the chemical forms of DNA characteristic of viruses and cells, including for example, simple and complex cells.
The term "patient" as used herein refers to a living human or non-human organism that is receiving medical care or that should receive medical care due to a disease. This includes persons with no defined illness who are being investigated for signs of pathology. Thus the methods and assays described herein are applicable to both, human and veterinary disease. The term "primer" as used herein, refers to an nucleic acid, whether occurring naturally as in a purified restriction digest or produced synthetically, which is capable of acting as a point of initiation of synthesis when placed under conditions in which synthesis of a primer extension product, which is complementary to a nucleic acid strand, is induced, i.e., in the presence of nucleotides and an inducing agent such as a DNA polymerase and at a suitable temperature and pH. The primer may be either single-stranded or double-stranded and must be sufficiently long to prime the synthesis of the desired extension product in the presence of the inducing agent. The exact length of the primer will depend upon many factors, including temperature, source of primer and the method used. Preferably, primers have a length of from about 15-100 bases, more preferably about 20-50, most preferably about 20-40 bases. The factors involved in determining the appropriate length of primer are readily known to one of ordinary skill in the art.
The term "sample" as used herein refers to a sample of bodily fluid or tissue obtained for the purpose of diagnosis, prognosis, or evaluation of a subject of interest, such as a patient. Preferred test samples include blood, serum, plasma, cerebrospinal fluid, urine, saliva, sputum, and pleural effusions. In addition, one of skill in the art would realize that some test samples would be more readily analyzed following a fractionation or purification procedure, for example, separation of whole blood into serum or plasma components. Thus, in a preferred embodiment of the invention the sample is selected from the group comprising a blood sample, a serum sample, a plasma sample, a cerebrospinal fluid sample, a saliva sample and a urine sample or an extract of any of the aforementioned samples as well as circulating tumour cells in blood or lymph, any tissue suspected to contain metastases as well as any source that may contain prostate tumour cells or parts thereof, including vesicles like exosomes, microvesicles, and others as well as free or protein-bound RNA molecules derived from prostate tumour cells. Preferably, the sample is a blood sample, most preferably a serum sample or a plasma sample. Importantly, urine (particularly after digital rectal examination) and ejaculate belong to the most preferable samples. Tissue samples may also be biopsy material or tissue samples obtained during operations.

Detailed description of the invention

The invention describes a method of diagnosis of prostate cancer. This method comprises analysing a sample taken from a patient and specifically determine the level of a biomarker or a combination of biomarkers in said patient sample. The result is then correlated to a threshold value and in the case it is above that threshold value, said patient sample is designated prostate cancer positive.
The invention relates to a group of sequences comprising SEQ ID NOs 1 to 42. The sequences are listed below. Due to space constraints, only the first 100 nucleotides are listed. The remaining part of the sequence can be found in the sequence protocol.

There are two types of sequences. First, some transcripts are known sequences that are already annotated in relevant databases. They are identified by their respective annotations. Second, new transcripts were identified that are not covered by known database. They are annotated as follows: XLOC_ followed by a number. The transcripts provide information about their genomic origin, but it may not necessarily represent the whole sequence of a given transcript. The sequences as detected might in some cases be longer or shorter. In the case of XLOC transcripts, if fragments are detected, these fragments may be as small as 150, 100, 50, 40, 30, 20, 10, 9, 8, 7, 6 or 5 nucleotides. The sequences are presented according to their relevance, with the most preferred sequences listed first.

Table 1: List of SEQ ID NOs. SEQ ID NOs 1 to 42 are listed together with the corresponding transcript and gene annotations. The first 100 nucleotides of each SEQ ID NO are shown.

SEQ ID	Transcript	Gene/transcript annotation	Sequence
1	1	Retro-RPL7
		Ensemble-ID gene: ENSG00000242899.1
		Locus (hg19): Chr3: 131,962,301-131, 963, 125 (-)
		Ensemble-ID transcript(s): ENST00000479738.1
2	2	XLOC_133897
		Ensemble-ID gene: none
		Locus (hg19): chr20: 45,377,600-45,380,719 (-)
		Ensemble-ID transcript(s): none
		Includes GenBank entries: AK128800.1, BC065739.1
3	3	AC144450.2
		Ensemble-ID gene: ENSG00000203635.2
		Locus (hg19): Chr2: 1,624,282-1,629,191 (-)
		Ensemble-ID transcript(s): ENST00000366424
4	4	RP11-279F6.1
		Ensemble-ID gene: ENSG00000245750.3
		Locus (hg19): Chr15 69,755,365-69,863,775 (+)
		Ensemble ID trancript: ENST00000558633
5	4	ENST00000558309
6	4	ENST00000560882
7	4	ENST00000559029
8	4	ENST00000558781
9	4	ENST00000498938
10	4	ENST00000559477
11	5	AC144450.1
		Ensemble-ID gene: ENSG00000228613.1
		Locus (hg19): Chr2: 1,550,437-1,623,885 (-)
		Ensemble-ID transcript(s): ENST00000438247.1
12	6	AC012531.25
		Ensemble-ID gene: ENSG00000260597.1
		Locus (hg19): Chr12: 54,413,694-54,416,373 (+)
		Ensemble-ID transcript(s): ENST00000562848.1
13	7	XLOC_068574
		Ensemble-ID gene: none
		Locus (hg19): chr14: 62,653,302-62,655,723 (+)
		Ensemble-ID transcript(s): none
14	8	RP1-207H1.3
		Ensemble-ID gene: ENSG00000231150.1
		Locus (hg19): chr6:38,890,805-38,920,875 (-)
		Ensemble ID transcript: ENST00000416948.1
15	8	ENST00000453417.1
16	8	ENST00000418399.1
17	9	XLOC_016724
		Ensemble-ID gene: none
		Locus (hg19): chr1: 177,827,793-177,841,757 (-)
18	10	RP11-314O13.1
		Ensemble-ID gene: ENSG00000260896
		Locus (hg19): Chr16: 80,862,632-80,926,492 (-)
		Ensemble ID transcript: ENST00000562231
19	10	ENST00000569356
20	10	ENST00000561519
21	10	ENST00000563626
22	11	XLOC_167596
		Ensemble-ID gene: none
		Locus (hg19): chr4: 67,964,836-67,975,652 (-)
23	12	XLOC_167595
		Ensemble-ID gene:
		Locus (hg19): chr4: 67,946,236-67,964,614 (-)
24	13	XLOC_156132
		Ensemble-ID gene: none
		Locus (hg19): chr3: 193,632,725-193,636,178 (-)
25	14	XLOC_156120
		Ensemble-ID gene: none
		Locus (hg19): chr3: 193,580,748-193,608,459 (-)
26	15	RP11-627G23.1
		Ensemble-ID gene: ENSG00000255545.3
		Locus (hg19): Chr11: 134,306,367-134,375,555 (+)
		Ensemble ID transcript: ENST00000533390
27	15	ENST00000531319
28	15	ENST00000528482
29	15	ENST00000532886
30	16	XLOC_047797
		Ensemble-ID gene: none
		Locus (hg19): chr12: 75,378,181-75,383,176 (+)
31	17	ANKRD34B
		Ensemble-ID gene: ENSG00000189127.3
		Locus (hg19): Chr5: 79,852,574-79,866,307 (-)
		Ensemble ID transcript: ENST00000338682
32	17	ENST00000508916
33	18	XLOC_243739
		Ensemble-ID gene: none
		Locus (hg19): chr9: 79,530,077-79,542,427 (-)
34	19	XLOC_198292
		Ensemble-ID gene: none
		Locus (hg19): chr6: 148,396,831-148,428,362 (+)
35	20	XLOC_068639
		Ensemble-ID gene: none
		Locus (hg19): chr14: 62,931,844-62,933,233 (+)
36	21	XLOC_172083
		Ensemble-ID gene: none
		Locus (hg19): chr4: 169,961,616-169,999,957 (-)
37	22	XLOC_172082
		Ensemble-ID gene: none
		Locus (hg19): chr4: 169,947,628-169,961,481 (-)
38	23	XLOC_112832
		Ensemble-ID gene: none
		Locus (hg19): chr2: 123,297,707-123,644,538 (+)
39	24	XLOC_243747
		Ensemble-ID gene: none
		Locus (hg19): chr9: 79,622,778-79,633,361 (-)
40	25	XLOC_243744
		Ensemble-ID gene: none
		Locus (hg19): chr9: 79,601,892-79,606,132 (-)
41	26	XLOC_126289
		Ensemble-ID gene: none
		Locus (hg19): chr2: 180, 988, 687-180,989,287 (-)
42	27	XLOC_172084
		Ensemble-ID gene: none
		Locus (hg19): chr4: 169,983,995-169,984,246 (-)

Several of these novel biomarkers significantly surpass the specificity and sensitivity of the biomarker PCA3, which is already used for prostate carcinoma (PCa) diagnosis. In the sequencing cohort, PCA3 proved to be associated with PCa, yet with a strong tendency to a decline in the high-risk group (Fig. 2).
In the test cohort, the novel biomarker Retro-RPL7 yielded an area under ROC curve (AUC) value of 0.935, compared to 0.851 for PCA3 (Fig. 3). A combination of our novel biomarkers Retro-RPL7 (SEQ ID NO. 1) and XLOC133897 (SEQ ID NO. 2) showed an AUC of 0.975 (Fig. 4).
Hence, in one aspect of the invention the invention relates to using pairs of the markers identified, in particular the air of SEQ ID NO. 1 and 2, or SEQ ID NO. 1 and 3 or 1 and 4 etc. Also, more than two markers may be used.
The experimental results demonstrate high specificity and sensitivity of the novel biomarkers for the detection of PCa.

Ideally, the level of transcript of the nucleic acids according to SEQ ID NO. 1 to 42 is compared to the level of transcript of a given other gene, such as a housekeeping gene. Suitable housekeeping genes are shown below in the section Housekeeping:

Table 2

	Housekeeping name
Housekeeping	GAPDH - Glyceraldehyde 3-phosphate dehydrogenase
	HPRT1 - hypoxanthine phosphoribosyltransferase 1
	HMBS - hydroxymethylbilane synthase
	TBP Tata box binding protein

The threshold values for the transcript herein are typically determined as follows:

Ideally the expression fold change between tumour and control tissue is larger than 1,5 fold, 2 fold, 3 fold, 4 fold and most preferably 5-fold. The p-value (T test) is < 2x10^-5.

The invention relates to the analysis of RNA biomarker levels. This can be accomplished by a number of methods, for instance PCR-based methods like quantitative reverse transcriptase PCR.
The invention relates to a method of qRT-PCR to specifically determine the level of biomarkers. Herein, the sample is mixed with a forward and a reverse primer that are specific for said sequence selected from the group of SEQ ID NO. 1 and 42 and amplification cycles are performed. Probes or primers are designed such that they hybridize under stringent conditions to said target sequence.
The invention also relates to a quantification of the level of the biomarker. After amplification, quantification is straightforward and can be accomplished by a number of methods. In the case when primers are used wherein at least one primer has a fluorescent dye attached, quantification is possible using the fluorescent signal from the dye. Various primer systems and dyes are available, such as SYBR green, Multiplex probes, TaqMan probes, molecular beacons and Scorpion primers. Other possible means of quantification are absorbance measurements at 260 and 280 nm.
Further, the invention comprises a kit for the screening, diagnosis of prostate cancer. The kit comprises at least one primer or probe, which is specific for one of the SEQ ID NOs. 1 to 42 and reagents for nucleic acid amplification and quantification or detection.
Superior to current diagnostic methods, the invention discloses a high specificity and sensitivity of our biomarkers in the diagnosis of prostate cancer.

EXAMPLES

Prostate carcinoma (PCa) patients who underwent radical prostatectomy (RPE) or were operated on prostate hyperplasia (BPH) at the University Hospital of Dresden were included in a retrospective clinical cohort aiming at identifying novel biomarkers for PCa. Approval of the local ethics committee as well as informed consent by the patients were obtained according to the legal regulations. For PCa patients, data on the clinical follow-up for at least 5 years were collected.

For identification of diagnostically relevant biomarkers by genome-wide RNA sequencing, prostate tissue samples from a cohort of 40 PCa patients and 8 BPH patients were used. Four PCa groups were defined based on staging according to Gleason (The Veteran's Administration Cooperative Urologic Research Group: histologic grading and clinical staging of prostatic carcinoma; in Tannenbaum, M. Urologic Pathology: The Prostate, Philadelphia: Lea and Febiger. Pp. 171-198) as well as the presence of metastases in the adjacent lymph nodes upon RPE (see Table 3).

Table 3: PCa cohort for genome-wide RNA sequencing: The control group (C) consisted of BPH samples. The very low risk (V) and low risk (L) groups comprised samples from patients graded with Gleason Score (GS)<7 and =7, respectively, all without lymph node metastases (pN0). The medium risk (M) group comprised GS<=7 cases exhibiting lymph node metastases (pN+), and the high risk (H) group tissues of GS>7. For the latter, pairs of tumour and tumour-free tissue samples obtained from the same patient were chosen.

Group	C	V	L	M	H
Gleason score	BPH	GS<7	GS=7	GS<=7	GS>7
lymph node metastasis	-	pN0	pN0	pN+	pN0		pN+
tissue	control	tumour	tumour	tumour	tumour	tumour-free	tumour	tumour-free
number of samples	8	8	8	8	8	8	8	8

Selected biomarker candidates were further validated by quantitative reverse-transcriptase real-time PCR (qRT-PCR) on a second cohort comprising 56 patients (16 control BPH samples, 40 tumour samples).

Prostate tissue samples

Prostate tissue samples were obtained from operations carried out at the Dept. of Urology of the University Hospital of Dresden and stored in liquid nitrogen at the Comprehensive Cancer Centre of Dresden University. Prostate tissue samples obtained from radical prostatectomies (RPEs) of prostate carcinoma (PCa) patients were divided into tumour and tumour-free samples. Prostate tissue samples from patients with benign prostate hyperplasia (BPH) were used as non-tumour controls. Patient consent was always given.
To verify the status of the samples and their tumour cell content, all samples were divided into series of cryosections. To this purpose, frozen tissue samples were embedded in Tissue-Tek OCT-compound (Sakura Finetek GmbH) and fixed on metal indenters by freezing. Cryosections were prepared using a cryomicrotome (Leica) equipped with a microtome blade C35 (FEATHER) cooled to -28°C. Every sample was cut into a total of 208 cryosections, 4 of which were HE-stained and evaluated by a pathologist with respect to their tumour cell content (Fig. 1). This yielded 3 stacks of consecutive cryosections, each of which was flanked by HE-stained sections. Only stacks that were flanked on either side by sections containing at least 60% or maximal 5% tumour cells were used as tumour or tumour-free samples, respectively. 50 cryosections of the stacks chosen were then subjected to RNA preparation.

RNA isolation

Total RNA was isolated using Qiazol and the miRNeasy Mini Kit on the QIAcube (all from Qiagen) with manual subsequent DNase I digestion. RNA concentration was determined using a Nanodrop 1000 (Peqlab). RNA integrity was verified on an Agilent Bioanalyzer 2100 (Agilent Technologies, Palo Alto, CA), and only RNA samples with an RNA-Integrity-Number (RIN) of at least 6 were further processed.

Genome-wide long-RNA next generation sequencing

1 µg of total RNA was depleted of ribosomal RNA using the Ribo-Zero rRNA Removal Kit (Epicentre). Sequencing libraries were prepared from 50 ng of rRNA-depleted RNA using ScriptSeq v2 RNA-Seq Library Preparation Kit (Epicentre). The di-tagged cDNA was purified using the Agencourt AMPure XP System Kit (Beckman Coulter). PCR was carried out through 10 cycles to incorporate index barcodes for sample multiplexing and amplify the cDNA libraries. The quality and concentration of the amplified libraries were determined using a DNA High Sensitivity Kit on an Agilent Bioanalyzer (Agilent Technologies). 4 ng each of 8 samples were pooled and size-selected on 2% agarose gels using agarose gel electrophoresis. The sample range between 150 bp and 600 bp was gel-excised and purified with the MinElute Gel Extraction Kit (Qiagen), according to manufacturer's instructions. The purified libraries were quantified on an Agilent Bioanalyzer using a DNA High Sensitivity Chip (Agilent Technologies). Cluster generation was performed using TruSeq PE Cluster Kits v3 (Illumina Inc.) in an Illumina cBOT instrument following the manufacturer's protocol. Every purified and size-selected library pool was then loaded onto an Illumina HiSeq2000 flow cell, distributing it among all lanes. Sequencing was performed on an Illumina HiSeq2000 sequencing machine (Illumina, Inc.). The details of the sequencing runs were as follows: paired-end sequencing strategy; 101 cycles for Read1, 7 cycles for index sequences, and 101 cycles for Read2.

Analysis of sequencing data: Raw data preparation

Raw sequencing data comprising base call files (BCL files) was processed with CASAVA v1.8.1 (Illumina) resulting in FASTQ files. FASTQ files contain for each clinical sample all sequenced RNA fragments, in the following referred to as "reads". Specific adapter sequences were removed by using cutadapt (http://code.google.com/p/cutadapt/).

Analysis of sequencing data: Genome mapping and transcript assembling:

Reads were mapped to the human genome (assembly hgl9) using segemehl v0.1.4-382 and TopHat v2.0.9. Novel transcripts, i.e. transcripts not annotated in Gencode v17, were assembled using Cufflinks v2.1.1 and Cuffmerge v2.1.1. All novel transcripts and all known Gencode v17 transcripts were combined into a comprehensive annotation set.

Analysis of sequencing data: Statistical analysis

Htseq-count v0.5.4p1 (http://www-huber.embl.de/users/anders/HTSeq/doc/count.html) was used to compute the read counts per transcript and gene that are contained in the comprehensive annotation set of novel and known transcripts. Differentially expressed transcripts and genes were identified using R and the Bioconductor libraries edgeR. Different RNA composition of the clinical samples was adjusted for by scaling library size for each sample (TMM method). A negative binomial log-linear model was fitted to the read counts for each transcript or gene, and coefficients distinct from zero identified by a likelihood ratio test. False discovery rate was controlled by Benjanimi-Hochberg adjustment.

Validation by quantitative real-time PCR

cDNA was synthesized from 100 ng total RNA using the High-Capacity Reverse transcription kit (Applied Biosystems) and random primers according to manufacturer's instructions. Subsequent PCR assays were run using 4 µl of the diluted cDNA. Quantitative real-time PCR was performed using custom- (Retro-RPL7, lincRNA AK129581) and pre-designed (all others) TaqMan Gene Expression Assays (Applied Biosystems) for three different housekeeping genes and several target transcripts (see Table 2) on an Applied Biosystems 7900HT Real-Time PCR System.

Table 2: IDs of the Applied Biosystems TaqMan Gene Expression Assays used for qRT-PCR validation.

	Housekeeping/Target name	TaqMan Assay ID
Housekeeping	GAPDH	Hs02758991_g1
	HPRT1	Hs02800695_m1
	HMBS	Hs00609293_g1
Target	Retro-RPL7	AJ70L28
	XLOC133897	Hs03664641_s1
	lincRNA LOC145837	Hs01388451_m1
	lincRNA AK129581	AJCSVRJ
	TMPRSS2-ERG	Hs03063375_ft
	RPL7	Hs02596927_g1
	PCA3	Hs01371939_g1

All samples were measured in triplicate and the means of these measurements were used for further calculations.

Statistical analysis of the RT-qPCR results

Data normalization was carried out against an unregulated housekeeping gene. For each target the appropriate housekeeping gene that exhibited comparable expression levels was selected for normalization. For relative quantification, changes in gene expression of each sample were analysed relative to the median expression of the control samples. All statistical analyses were carried out using R statistical software.
The log2-transformed relative expression levels of the biomarkers were compared between tumour and control samples employing Student's t-test. Receiver-operating characteristic (ROC) curves, representing a measure of diagnostic power of each marker by the area under the curve (AUC), were calculated using the package pROC. Furthermore, logistic regression models were calculated for multiple-gene signatures using the package epicalc.

Results

The transcriptomes of 40 PCa tumour samples and 16 tumour-free samples obtained upon RPE and 8 BPH prostate tissue samples as benign, non-tumour controls were analysed using strand-specific, paired-end long RNA next generation sequencing (NGS). Approximately 150 cryosections per sample in at least three segments were prepared, aiming at an optimal data quality and robustness of the analysis. Upon pathological evaluation, only segments satisfying a maximal and minimal tumour cell count of 60% and 5% in tumour and tumour free samples, respectively, were retained for further analysis. The transcriptome sequencing (RNAseq) approach aimed at a comprehensive identification and quantification of RNAs expressed in normal or cancer prostate tissue. All classes of coding and long non-coding transcripts independent of polyadenylation status were sequenced. Large input masses of RNA were used to ensure high library complexity. Furthermore, on average 200 M paired-end reads 2 x 100 nt per library were sequenced, enabling the assembly of novel lowly expressed transcripts due to high coverage. This approach outperformed most comparable published studies that analyzed larger numbers of samples. In total, approx. 3000 novel transcripts that did not show an exonic overlap with transcripts annotated in Gencode v17 were assembled. At a false discovery rate of 0.01, 6442 differentially expressed genes across all contrasts were observed. Numbers of differentially expressed genes for specific contrasts are given in Table 4.

Table 4: Number of differentially expressed genes for diverse contrasts and Gencode biotypes.

Contrast	Total	Protein coding	lincRNA	Antisense	Sense-intronic	Pseudogene	Novel transcript	Non-protein coding
Tumour vs. Control	5615	3882	116	96	13	456	847	1733
Tumour Gleason > 7 vs. control	2677	1812	73	40	4	88	552	865
Tumour high and medium vs. Tumour low and very low	138	51	3	2	0	7	72	87
Tumour Gleason = 7 vs. Tumour Gleason < 7	12	6	0	1	0	0	5	6
Tumour Gleason > 7 vs. Tumour Gleason = 7	14	7	0	0	0	1	6	7

The results successfully reproduced the majority of transcripts previously reported to be differentially expressed between prostate tumour and normal tissue. In addition, a number of novel PCa-associated transcripts were identified, which can be used to develop assays for the diagnosis of PCa. Selected most promising transcripts were validated in a test cohort of PCA tumour and BPH control samples by qRT-PCR.
Several of these novel biomarker candidates significantly surpass the specificity and sensitivity of the biomarker PCA3, which is already used for PCa diagnosis. In the sequencing cohort, PCA3 proved to be clearly associated with PCa, yet with a strong tendency to a decline in the high-risk group (Fig. 2).
In the test cohort, the novel biomarker Retro-RPL7 yielded an area under ROC curve (AUC) value of 0.935, compared to 0.851 for PCA3 (Fig. 3). A combination of our novel biomarkers Retro-RPL7 and XLOC133897 showed an AUC of 0.975 (Fig. 4).
The experimental results demonstrate high specificity and sensitivity of the novel biomarkers for the detection of PCa. Therefore, assays can be set up based on the measurement of these newly discovered biomarkers alone or in combination (or in combination with already known markers) in all sources that may contain prostate tumour cells or parts thereof (including vesicles like exosomes, microvesicles, and others as well as free or protein-bound RNA molecules deriving from prostate tumour cells) to be used for the diagnosis of PCa. These sources include (but are not limited to) prostate tissue, biopsy material, lymph nodes, urine, ejaculate, blood, blood serum, blood plasma, circulating tumour cells in blood or lymph, as well as any tissue suspected to contain PCa metastases. Measurement of our RNA biomarkers can be done by any method suited to specifically estimate RNA levels, e.g. PCR-based methods like qRT-PCR. The assays can be applied for early diagnosis (screening) of PCa, for predicting the aggressiveness of the tumours (prognosis), and/or for aiding the choice of therapy.
Diagnostic assays based on these biomarkers may therefore dramatically decrease the high false-positive rates of current assays and thereby help to avoid unnecessary invasive prostate biopsies.

Figure captions

Fig.1: Verification of tissue sample quality: to ensure tumour cell content of the tissue samples, cryosections were prepared from the frozen samples as shown. HE: hematoxylin/eosin; IHC: immunohistochemistry. Verification of tissue sample quality: cryosections à 4µm were prepared from the frozen samples as shown for HE staining (to ensure tumour cell content of the tissue samples), for RNA and DNA isolation and for IHC. HE: hematoxylin/eosin; IHC: immunohistochemistry.
Fig.2: Box-blot of RNA-seq data for transcript PCA3. Results from RNA sequencing of the retrospective PCa cohort comprising 8 prostate tissue samples from benign prostate hyperplasia as a control (C), 8 PCa tumour samples each of groups V (very low risk; Gleason <7, pN0), L (low risk; Gleason =7, pN0), and M (medium risk; Gleason <=7, pN+),as well as 16 pairs of tumour and tumour-free tissue samples from group H (high risk; Gleason >7).
Fig.3: ROC curves of Retro-RPL7 and PCA3.
Fig.4: Logit model based on a two-transcript signature. ROC curve is shown for the multivariate logit model comprising Retro-RPL7 and XLOC133897 (AUC=0.975).

Claims

A method for the diagnosis of prostate cancer, comprising the steps of
a) analyzing a sample of a patient for the presence and/or level of nucleic acids selected from the group of SEQ ID NOs 1 to 42,

b) wherein, if at least one of said nucleic acids is present and/or the level of at least one of said nucleic acid is above a threshold value, said sample is designated as prostate cancer positive.
A method according to claim 1, wherein the nucleic acids are selected from the group of SEQ ID NOs. 1 to 42.
A method according to claims 1 or 2, wherein the sample is from one of the following sources: prostate tissue, biopsy material, lymph nodes, urine, ejaculate, blood, blood serum, blood plasma, circulating tumour cells in blood or lymph, any tissue suspected to contain metastases as well as any source that may contain prostate tumour cells or parts thereof, including vesicles like exosomes, micro vesicles, and others as well as free or protein-bound RNA molecules derived from prostate tumour cells.
A method according to claims 1 to 3, wherein analysis of said nucleic acid selected from the group of SEQ ID NO. 1 to 42 is done by any method suited to specifically determine nucleic acid levels, for instance PCR-based methods such as qRT-PCR.
A method according to claim 4, wherein the method to specifically determine nucleic acid levels employs the technique of qRT-PCR, comprising the steps of
a) mixing a sample with a forward and a reverse primer that are specific for said sequence as selected from SEQ ID NOs 1 to 42 and

b) performing reverse transcriptase quantitative PCR.
A method according to any of the preceding claims, wherein at least one of the primers has a fluorescent dye attached.
A probe or primer that hybridizes under stringent conditions to one of the said nucleic acids according to SEQ ID NO. 1 to 42.
A probe or primer according to claim 8, wherein the probe or primer is a nucleic acid and it is about 10 to 100 nt in length.
A probe or primer according to claims 8 to 9, wherein the probe or primer is specific for one of said sequences as selected from the group of SEQ ID NO. 1 to 42 of claim 1.
A nucleic acid from the group of SEQ ID NO. 1 to SEQ ID NO. 42 or the reverse complement thereof, or a nucleic acid that shares preferably at least 85%, 90%, 95% or 99% sequence identity with a nucleic acid according to any one of the nucleic acids according to SEQ ID NO. 1 to 42.
Use of a nucleic acid selected from the group of SEQ ID NO. 1 to 42 for the diagnosis of prostate cancer.
A kit for the diagnosis of prostate cancer comprising at least one probe or primer according to claims 8 to 10 and reagents for nucleic acid amplification and quantification or detection.